Tablebase 1 - Botvinnik 0

The diagram and text on the left is from Botvinnik's 'Soviet Chess Championship, 1941' (p.154). It shows the position after 84.Kf2-f1 in Smyslov - Lilienthal, from rd.16 of the event played in March-April 1941.

Note the concluding sentence: 'The given position is won, since the White King cannot get into the draw area.' Botvinnik continued, 'Lilienthal did not have a very easy task when studying Troitsky's analyses. However, that is no excuse for his further weak play. In general, this endgame is a rare occurrence in practical play. However, if my memory does not betray me, Lilienthal had encountered this very endgame (irony of fate!) twice previously, and on neither occasion could he discover the way to victory.'

In fact, a five piece endgame tablebase shows that the position after Smyslov's 84.Kf1 is a draw. Botvinnik gave 84...Nc2 85.Kg2 Ke3 86.Kg3, with analysis through move 97 to show how Black wins. The problem is that White has the paradoxical 86.Kh3, moving to the side of the board. This draws after 86...Kf3 (86...Ne1 87.Kg3) 87.Kh4 Kf4 88.Kh5 Kf5 89.Kh4, when Black can't make progress.

Lilienthal tried instead 84...Ne6. After 85.Kf2 Neg5 86.Kg3 Ke3 87.Kg4 Ke4 88.Kg3 Nf3, Smyslov continued to play perfectly and held his well earned theoretical draw. After Black's 96th move, Botvinnik wrote that it 'also should not have led to a draw', but he was wrong there as well.

I don't have access to Troitsky's original analysis, so I'm not sure what he overlooked. In the variation given above, after 86.Kh3 through 88...Kf5, I tried the Knight on squares other than c2 and found several other positions where White draws even though the King is outside the 'black line'. Troitsky's analysis was not correct and Botvinnik was wrong to condemn Lilienthal for 'weak play'.